RuView: WiFi-based spatial sensing on ESP32
Built it. And it actually works. Not currently entirely sure how. A WiFi sensing platform that turns Channel State Information (CSI) from low-cost ESP32 sensors into spatial intelligence: presence and occupancy detection through walls, contactless breathing and heart-rate measurement, activity recognition, RF fingerprinting, and overnight sleep monitoring.
What it senses
Presence and occupancy through walls. Breathing rate and heart rate, contactless, while sleeping or sitting. Walking, sitting, gestures, and falls from temporal CSI patterns. RF fingerprinting that identifies rooms, detects moved furniture, and spots new objects. Sleep stage classification and apnea screening.
Hardware and limits
Requires ESP32 with sufficient processing power for CSI DSP. ESP32-C3 and the original single-core ESP32 are not supported. Single-node deployments have limited spatial resolution. PCK@20 sits around 2.5% with proxy labels; the camera-supervised training pipeline targets 35%+ but data collection and evaluation are still pending.
Where to find it
github.com/ruvnet/RuView. Beta software under active development.